Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana

G Model

ECOHYD 132 1–14 Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Ecohydrology & Hydrobiology journal homepage: www.elsevier.com/locate/ecohyd

1 2 3 4

Original Research Article

Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana Mosepele a,*, J. Kolding b, T. Bokhutlo c

5

Q1 K.

6 7 8

Q2 a University of Botswana, Okavango Research Institute, Private Bag 285, Maun, Botswana b c

University of Bergen, Department of Biology, High Technology Centre, N-5020 Bergen, Norway Botswana International University of Science and Technology, College of Science, Private Bag 16, Palapye, Botswana

A R T I C L E I N F O

A B S T R A C T

Article history: Received 6 November 2016 Accepted 31 January 2017 Available online xxx

Tropical floodplain fish populations fluctuate at temporal scales and understanding the variability in these systems will contribute to comprehensive management of these resources. Therefore, the aim of this study was to assess the dynamics of a floodplain fish assemblage. Data were collected using standard methods between 1999 and 2009 from the Delta’s panhandle. Various analytical tools (e.g. CCA, SIMPER, ANOVA, etc.) were used to assess fish assemblage dynamics at seasonal and annual scales. ANOVA and cluster analyses showed that the fish assemblage underwent significant changes along the seasonal hydrograph, while %IRI revealed that the fish assemblage was dominated by Clarias gariepinus, Schilbe intermedius and Hydrocynus vittatus respectively. These species, including Clarias ngamensis and Marcusenius altisambesi, contributed more than 50% to variations in fish assemblage structure along the seasonal hydrograph (based on SIMPER analysis). Furthermore, CCA revealed a significant (p = 0.004) association between environmental factors and fish assemblage structure. CCA analyses also showed that spawning for different species is associated with various environmental factors. Annually, results showed that C. gariepinus dominated the fish assemblage during poor flood years while S. intermedius dominated during high flood years. DCA analyses showed that the hydrological gradient had a significant effect on fish assemblage structure at an annual scale, while SIMPER analyses established significant variations in fish assemblage structure among years characterized by different hydrological features. One major conclusion we made was that fish assemblages are stochastically different at an annual scale. This study contributes knowledge to floodplain fish ecology and thus enhances fisheries management. ß 2017 European Regional Centre for Ecohydrology of the Polish Academy of Sciences. Published by Elsevier Sp. z o.o. All rights reserved.

Keywords: Inter-annual variability Seasonal variability Flood pulse Floodplain fish ecology Okavango Delta

9 10

1. Introduction

11 12

In tropical flood plains fish biomass is directly related to seasonal flooding (Lowe-McConnell, 1987; Welcomme

* Corresponding author. Fax: +267 686 1835. E-mail addresses: [email protected], [email protected] (K. Mosepele).

et al., 2006). The underlying dynamic relationships are encapsulated in the flood pulse concept (Junk et al., 1989) which integrates the interactions between hydrological and ecological processes (Tockner et al., 2000). The flood pulse enhances biological productivity and maintains species diversity (Bayley, 1995) and seasonal fish migrations caused by the flood pulse facilitate the transmission of energy from the terrestrial environment to the aquatic system (Junk et al., 1989). Fish growth, mortality and

http://dx.doi.org/10.1016/j.ecohyd.2017.01.005 1642-3593/ß 2017 European Regional Centre for Ecohydrology of the Polish Academy of Sciences. Published by Elsevier Sp. z o.o. All rights reserved.

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

13 14 15 16 17 18 19 20 21

G Model

ECOHYD 132 1–14 2

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

breeding are directly related to the flood strength (Halls et al., 1999; de Graaf, 2003). Despite this highly dynamic relationship between climate driven hydrology and biological productivity, most fisheries management in floodplain systems is based on managing internal drivers (e.g. fishing effort) only (Welcomme, 2007) which is largely based on steady state assumptions. Therefore, there is a strong need to understand the effects of environmental variability on floodplain fish assemblages to inform fisheries management. A suite of factors is responsible for spatio-temporal fluctuations in the floodplain fish assemblage structure. The seasonally inundated floodplain, lagoons and riparian zones are the most important habitats that regulate fish productivity, community structure, and diversity (Ward and Tockner, 2001). Fish species diversity within floodplain communities is typically highest at high floods and is lowest at low flood levels when there is low connectivity (Ward and Tockner, 2001). The aim of this study was to explore Junk et al’s. (1989) flood pulse concept in the Okavango Delta by exploring the presence of the flood pulse in the Delta’s fish assemblage.

44

2. Materials and methods

45

2.1. Study area

46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Okavango Delta (Fig. 1) is one of the world’s largest inland deltas (Ramberg et al., 2006a). While local rainfall has a localized impact (Wolski et al., 2005), the Delta’s hydrology is driven by annual flooding from Angola (Wolski and Savenije, 2006) with a strong inter-annual variability (Fig. 1). Discharge into the Delta’s northern panhandle peaks in April (Fig. 1) and is generally out of phase with the rainy season in the Delta (Wolski and Savenije, 2006). The peak flood pulses through the entire system and usually takes 1–2 months from Mohembo to Seronga and another 2–3 months to reach the distal end of the Delta in Maun (Wolski et al., 2005). There are 71 fish species in the Delta (Ramberg et al., 2006b) distributed heterogeneously throughout the system (Mosepele et al., 2009).

61

2.2. Data collection

62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77

Fish data: Experimental fish data were collected between 1999 and 2009 (there was no sampling in 2003) at Ngarange and Seronga (Fig. 1) sampling stations. Two types of experimental nets were used: (i) multifilament, multi-mesh, nets made up of 9, 10 m long panels of mesh sizes 22–150 mm stretched; (ii) multifilament, multi-mesh nets made up of 5.5 m long panels of mesh sizes 50–125 mm stretched mesh. Sampling was done at each station 2–3 days monthly. Nets were set for approximately 12 h overnight and 10 h during the day (to account for diurnal variations in fish movements). Nets were set along the margins of the main channel and in a lagoon in each sampling station. The main channel has a sandy bed, fringed by papyrus (Cyperus papyrus) and reeds (Phragmites australis) rooted in mud rich peat, with water flow velocity ranging between 0.4 and 0.8 ms 1 (McCarthy

et al., 1998; Wolski et al., 2006). Lagoons are seasonally connected to the min channel by narrow channels (Gondwe and Masamba, 2013) and fringed by papyrus, reeds and typha beds (Smith, 1976) with relatively sluggish water velocity (Mendelsohn et al., 2010). Catches from each panel were separated and recorded separately. The sampling regime and data treatment are described in Mosepele (2000). Maturity stages were based on Nikolsky’s (1969) six stage key where stage 5 is ripe running (spawning). The data from the two nets were harmonised by using only data from mesh sizes 49–125 mm. This amounted to 57,222 fish records that were used in this study.

78 79 80 81 82 83 84 85 86 87 88 89 90

2.3. Data analysis

91

General statistics: Multiple linear step-wise regression in STATISITCA (Version 6.0, StatSoft) was used to determine the strength and significance of relationships between variables (with a significance level of 0.05, except a few cases of 0.1). The relative strength of independent variables was determined by the magnitude of the p value (Zeug and Winemiller, 2007). ANOVA was used to test for the level of differences among variables along temporal scales. Univariate analysis: Fish indices (in either numbers/set or grams/set), spawning, index of relative abundance (%IRI), and mean length were calculated in Pasgear (Kolding and Asmund, 2010). Spawning season was defined as a period when a minimum of 5% of the adult population was spawning. The %IRI is considered a good measure of abundance, as it combines numbers, weight and frequency of occurrence (Hart et al., 2002). Fish abundance data were clustered into four seasonal discharge stages (increasing, peak, decreasing and minimum). The assemblage stability was assessed using the coefficient of variation (CV) (Grossman et al., 1990; Oberdoff and Porcher, 1992). Scaling population variation by the mean permits comparison of populations with different mean abundances which makes it less ambiguous than other metrics (Grossman et al., 1990). CV percentage values were classified into equal quartiles as stable, moderately stable, moderately fluctuating and fluctuating (Freeman et al., 1988). Multivariate analysis: Cluster analysis in PRIMER 6 (Clarke and Gorley, 2001) was used to establish assemblage patterns (Minns, 1989). All data for analysis in Primer were standardized, and then square-root transformed before using Bray–Curtis similarity analysis. Spatio-temporal differences in fish assemblage structure were assessed using SIMPER analysis (Rayner et al., 2015). SIMPER scores were then plotted to explore patterns over temporal scales and hydrological variables. Environmental effects on the Delta’s fish species assemblage and spawning behavior were assessed by direct gradient analysis using canonical correspondence analysis (CCA) (ter Braak, 1986) implemented in PcORD v6 (McCune and Mefford, 2006). Data were log transformed (log(x + 1)) to minimize the range and skew of distributions (Cantu and Winemiller, 1997). The Euclidean distance was used as the dissimilarity measure while p was estimated by a Monte

92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

3

Fig. 1. (a) Map of the Okavango Delta highlighting the panhandle. This figure also illustrates environmental variability in the Delta through (b) inter-annual variability in discharge, (c) inter-annual variability in discharge and flooded area during the study period, and seasonal variability in discharge and (d) flooded area, (e) temperature, (f) pH, (g) TDS, and (h) DOC. (Note: Graphs D–H present mean values over the study period).

137 138 139 140 141 142 143 144

Carlo test using 999 permutations (McCune and Mefford, 1999). We also used DCA/DECORANA to search for patterns in assemblage shifts (Ibarra and Stewart, 1989) using Shannon’s diversity index. Regression analysis was used to explore the strength and nature of relationship between diversity and hydrological variables. Only axis scores from DCA analysis that had significant relationships were included in the analysis.

3. Results

145

3.1. General observations

146

Clarias gariepinus, Schilbe intermedius and Hydrocynus vittatus (all predators) dominated the fish assemblage at both seasonal and annual scales (Fig. 2). As summarized in Table 1, C. gariepinus was the most abundant species by relative weight (443 g/set) and S. intermedius by relative

147 148 149 150 151

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 4

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Fig. 2. Intra and inter annual variations in biodiversity in the Delta within the study period where the black lines show (A) intra-annual and (B) inter-annual variability in mean discharge.

152 153 154 155 156 157

numbers (0.970 fish/set) over the study period. The top three largest species (ML) in the composition were C. gariepinus (52.72 cm), C. ngamensis (44.85 cm) and H. vittatus (40.14 cm) respectively. Conversely, Petrocephalus okavangensis (8.53 cm), Barbus poechii (9.09 cm) and Brycinus lateralis (10.12 cm) were the smallest sized

species. The feeding guilds with the largest frequencies in the assemblage were insectivores (approximately 28%), omnivores (approximately 23%) and predators/carnivores (approximately 16%) respectfully. As summarized in Table 1, populations of the top three species had different stability dynamics (based on CV

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

158 159 160 161 162 163

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

5

Table 1 Species composition and relative abundance from the study area based on data collected between 1999 and 2009. Letters in () indicate the feeding guild of each species where superscript M is Mosepele et al. (2012), MB is Merron and Bruton (1988). Fecundity estimates are also derived from various sources where superscript M is Mosepele et al. (2012), MB is Merron and Bruton (1988) W is van der Waal (1985), S is Skelton (2001) and MM is Merron and Mann (1995). Family

Species

Species code

%IRI

No/set (density)

g/set (relative abundance)

Mean length (cm)

Fecundity

Mean weight (g)

CV (seasonal)

CV (annual)

Bagridae

Parauchenoglanis ngamensis(O)MB Zaireichthys cf conspicuus(I)MB

30 40

0.002 0

0.004 0

0.338 0.052

19.694 36.000

1000MB 16S

78.954 390.000

91 346

177 283

Characidae

Hydrocynus vittatus(C)MB Brycinus lateralis(OPA)M

3 31

10.617 0.002

0.244 0.008

166.891 0.076

40.144 10.120

150000MB 10000MB

685.169 9.454

87 175

44 267

Cichlidae

Serranochromis macrocephalus(P/C)MB Serranochromis angusticeps(P/C)MB Oreochromis andersonii(OPP)M Coptodon rendalli(OPP)M Tilapia sparrmanii(O)MB Serranochromis altus(P) Serranochromis robustus(P/C)MB Sargochromis carlottae(O)MB Oreochromis macrochir(D)MB Sargochromis codringtonii(M)MB Serranochromis thumbergi(P/C)MB Sargochromis giardi (O)MB Sargochromis greenwoodii(M)MB Pharyngochromis acuticeps(P/C)MB Tilapia ruweti(D)MB Hemichromis elongatus(O)MB Pseudocrenilabrus philander(P/C)MB Serranchromis longimanus(P/C)MB

9 10 14 15 17 18 19 20 22 23 24 25 28 35 36 37 38 39

0.451 0.432 0.187 0.178 0.142 0.074 0.061 0.057 0.037 0.037 0.025 0.019 0.002 0 0 0 0 0

0.056 0.058 0.031 0.037 0.052 0.017 0.015 0.022 0.014 0.018 0.014 0.010 0.004 0.001 0.001 0 0 0

17.713 17.144 15.729 11.262 4.439 10.824 8.904 3.479 5.061 3.185 2.193 4.227 0.875 0.097 0.114 0.022 0.012 0.037

26.527 26.653 29.177 23.909 14.972 33.398 31.733 19.489 25.001 20.817 21.125 25.869 23.007 18.867 19.775 15.567 11.933 20.250

225MB 5000MB 2000MB 7000MB 800MB – 1500MB 500MB 572W 700MB 300MB 1200MB – – 400MB 500MB 400MB

314.381 297.207 502.267 300.526 86.143 652.242 583.591 157.526 356.771 178.920 157.578 421.083 218.006 120.833 212.750 54.667 30.000 137.500

86 93 117 118 144 95 105 81 136 75 76 96 183 233 280 205 195 287

81 121 90 55 148 102 72 120 101 88 180 180 165 278 283 283 271 283

Clariidae

Clarias gariepinus(OPA) Clarias ngamensis(OPP)M Clarias theodorae(O)MB Catfishes

1 4 27 –

39.111 8.568 0.004 0

0.376 0.228 0.005 0.001

443.903 163.383 2.048 0.719

52.717 44.850 37.383 34.000

200000MB 200000MB – –

1180.372 715.592 425.083 488.073

44 100 203 346

35 44 159 –

Cyprinidae

Labeo lunatus(D)MB Barbus poechii(I)MB

32 34

0.001 0

0.003 0.005

0.833 0.032

31.225 9.085

– –

311.100 7.002

133 229

106 276

Hepsetidae

Hepsetus cuvieri (CPF)M

7

2.429

0.166

47.59

30.733

8000MB

287.230

84

93

Synodontis nigromaculatus(O) Synodontis thamalakanensis Synodontis vanderwaali Synodontis leopardinus(O)MB Synodontis woosnami(O)MB Synodontis macrostigma Squeakers (Synodontis sp.) Synodontis macrostoma

5 8 11 13 16 21 26 29

5.995 0.538 0.420 0.196 0.163 0.048 0.008 0.002

0.427 0.093 0.088 0.054 0.052 0.027 0.016 0.005

31.672 5.835 6.416 4.089 4.049 1.557 0.787 0.314

20.101 18.069 18.953 19.687 18.023 16.024 15.631 16.222

31333W

40334W 40334W 45000W – –

74.116 62.824 72.742 76.009 77.379 57.322 50.250 57.293

67 124 72 82 135 123 136 287

31 42 94 109 48 143 110 157

Mormyridae

Marcusenius altisambesi(OPA)M Mormyrus lacerda(I)MB Petrocephalus okavangensis(I)MB

6 12 33

5.537 0.365 0.001

0.534 0.052 0.005

36.672 18.420 0.026

18.926 33.361 8.531

5000MB 7000S –

68.641 355.636 5.494

90 108 224

86 110 282

Schilbeidae

Schilbe intermedius(OPA)M

2

24.293

0.970

106.137

23.303

27000MM

109.432

56

52

Mochokidae

M

MB

Note: 1 net set = 1 net set for 12 h overnight; IRI = index of relative importance; O = Omnivore; I = Insectivore; D = Detritivore; M = Molluscivore; P/C = Predator/Carnivore; OPA = Omnivore with preference for animals; OPP = Omnivore with preference for plants; CPF = Carnivore with preference for fish.

164 165 166 167 168 169 170 171 172 173 174 175

values) at intra and inter annual scales. Intra-annually, C. gariepinus population was moderately stable (44); S. intermedius moderately fluctuating (56), while H. vittatus was highly fluctuating (87). Inter-annually, C. gariepinus (35) and H. vittatus (44) populations were moderately stable, while S. intermedius (52) were moderately fluctuating. Populations of the three smallest species were fluctuating at both seasonal and annual scales. Generally, 90% of the populations were fluctuating at the seasonal scale compared to 78% at the annual scale. In general, more species were moderately stable (15%) at the annual than at the seasonal scale (2%).

3.2. Intra-annual variability

176

The fish assemblage showed strong seasonal variability while regression analysis (Fig. 3) revealed that flooded area had a stronger effect on assemblage dynamics than discharge. Generally, changes in discharge explained 41% of the observed variability in fish assemblage patterns while flooded area explained 54% of the variability (Fig. 3). SIMPER analysis (Fig. 4) showed varying similarity among the fish assemblages at a seasonal scale and also along the two hydrological gradients (i.e. discharge and flooded area). Similarity was highest between September and

177 178 179 180 181 182 183 184 185 186

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 6

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Fig. 3. (i) Results of seasonal DCA analysis (diversity) scores against (A) inundated area (ANOVA: DF = 11; F = 11.67; p = 0.01; R2 = 0.54; Beta = 0.73), (B) discharge (Regression: DF = 11; F = 7.01; p = 0.02; R2 = 0.41; Beta = 0.64), and annual DCA analysis (diversity) scores against (C) discharge (Regression: DF = 7; F = 4.91; p = 0.07; R2 = 0.45; Beta = 0.67) and (D) flooded area (Regression: DF = 7; F = 4.98; p = 0.07; R2 = 0.45; Beta = 0.67) from the study area. Also illustrated is (E) inter-annual variability in relative abundance and discharge (Regression: DF = 7; F = 10.43; p = 0.02; R2 = 0.63; Beta = 0.80).

187 188 189 190 191 192 193 194 195 196 197 198 199 200

October, which coincides with maximum inundated area (see Fig. 1D), and lowest between April and August (increasing water volume). Conversely, similarity in the fish assemblage was lowest between peak vs decreasing discharge stage and highest at peak discharge stage (Fig. 4E). Moreover, ANOVA analysis (Table 2) revealed that assemblage stability (CV) changed significantly during the season. Cluster analysis (Fig. 5A) confirmed the strong intra annual variability into four hydrological groups (increasing, decreasing, minimum and peak discharge). H. vittatus, S. intermedius, C. gariepinus, C. ngamensis and Marcusenius altisambesi were the key species which contributed at least 50% to fish assemblage shifts (SIMPER analysis) within each hydrological cluster. ANOVA

(Table 2) showed that these species underwent significant seasonal changes in relative abundance. Regression analysis (Table 3) also revealed a significant relationship between hydrological variables (discharge and inundated area) and selected variables of these key fish species. Furthermore, C. gariepinus dominated the fish assemblage at peak inundation, while H. vittatus dominated the fish assemblage at peak discharge (Fig. 5A). CCA analysis (Table 4) revealed a strong (Pearson correlation: Axis 1 =0.993) and significant (Monte Carlo randomization: p = 0.004) correlation between fish species distribution in the study area. Overall, pH had a minimal effect on species distribution while temperature and discharge had a major effect (Fig. 6A). According to

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

201 202 203 204 205 206 207 208 209 210 211 212 213 214

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

7

Fig. 4. SIMPER analysis illustrating intra (A–C) and inter (D–F) annual variability in fish species assemblages, where the trend line illustrates the strength of the relationship between the species assemblages. Table 2 Summary of ANOVA analysis illustrating changes in relative abundance of selected fish species and assemblage stability (CV) along the seasonal (discharge) hydrograph (M = minimum; D = decreasing; I = increasing; P = peak). Significant values (at 95% confidence level) are in bold.

215 216 217 218 219 220 221 222 223

Index

Test

H. vittatus

C. gariepinus

S. intermedius

C. ngamensis

M. altisambesi

Relative abundance (g/set)

M vs. D M vs. I M vs. P D vs. I D vs. P P vs. I

0.018 0.023 0.085 0 0.001 0.770

0.064 0 0.004 0.043 0.095 0.006

0.665 0.009 0.034 0.004 0.022 0.135

0.023 0 0.004 0.135 0.077 0.021

0.488 0.049 0.105 0.087 0.101 0.329

Assemblage stability (CV)

M vs. D M vs. I M vs. P D vs. I D vs. P P vs. I

0.078 0.186 0.018 0.014 0.253 0.006

Table 4, discharge and DOC were associated with Axis 1, temperature with Axis 2, while area and DO were associated with Axis 3. Fig. 6A shows that H. vittatus was associated with a Discharge: DOC gradient during the peak discharge season. Several cichlid species and some Synodontis species were linked with a Temperature: DO gradient. Oreochromis macrochir was associated with an increasing discharge period, while Synodontis leopardinus, Synodontis woosnami and C. rendalli were associated with a

decreasing area period. Serranochromis angusticeps and Hepsetus cuvieri were associated with a pH: Area gradient. All three axes in the ordination cumulatively explained approximately 78% of the total variation in the fish assemblage. Spawning: CCA (Fig. 6B) revealed a strong (Pearson correlation: Axis 1 = 0.99) and significant (Monte Carlo randomization; p = 0.00) association between spawning of selected fish species and the environment (Table 4).

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

224 225 226 227 228 229 230 231 232

G Model

ECOHYD 132 1–14 8

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Fig. 5. CCA biplot illustrating the environmental effects on fish assemblage (A) distribution and (B) spawning behavior in the study area.

233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248

Cumulatively, the three axes explained approximately 69% of the observed variance in spawning behavior. Spawning by different fish species was associated with different environmental gradients (Fig. 6B, Table 4). Spawning for some species like H. vittatus and some cichlids (e.g. Serranochromis altus, O. macrochir, etc.) was associated with Axis 1, Axis 2 for some cichlids (e.g. Sargochromis giardi, Serranochromis macrocephalus, etc.) while spawning for species like C. gariepinus and S. intermedius was associated with Axis 3 (Table 4). Regression analysis (Table 5) revealed that discharge and temperature (each) had significant relationship with 24% of the species, mean flooded area with 36% of the species, and DO with 8% of the species. Most of the species had extended, but different, spawning seasons (Table 5). Approximately 38% of the cichlids in the study start spawning between June and

August, about 67% of Synodontis spp. start in October; approximately 67% of Silurids start in November, while about 67% of Mormyrids start in December. Approximately 52% of the species were spawning by September. Spawning season for 32% of the species ended in January (three months before peak discharge), and season for 36% of the species ended in March (a month before peak discharge). Peak spawning period for B. poechii and B. lateralis coincided with maximum flooded area (see Table 5).

249 250 251 252 253 254 255 256 257

3.3. Inter-annual variability

258

The fish assemblage structure underwent inter-annual variability (Fig. 2), while DCA analysis revealed that variability in the fish assemblage structure was driven by the inter-play of discharge and flooded area (Fig. 3C and D). Generally, DCA/DECORANA analysis showed that hydrology accounted for inter-annual variability in fish species distribution (Fig. 7). SIMPER analysis revealed that similarity in fish assemblage shifted along a hydrological gradient (Fig. 4C and D) and along a time scale (Fig. 4F). Similarity was highest during low discharge (Fig. 4C) and inundated area (Fig. 4D) years and least during high discharge and inundated area years. C. gariepinus and S. intermedius were the two main species that contributed most to these assemblage dynamics. Generally, the highest similarity in fish assemblages was observed between 2001 vs. 2002 while the lowest similarity was observed between 2007 vs. 2009 (Fig. 4F). Regression analysis (Table 6) revealed that hydrology was a driver of change among the key fish species dynamics. Cluster analysis (Fig. 6B) grouped the fish assemblage into five groups along a hydrological scale. SIMPER showed that the top three species contributing to these assemblage shifts were C. gariepinus, H. vittatus and S. intermedius. Furthermore, C. gariepinus dominated the fish assemblage during years of low flooded area while S. intermedius dominated during high discharge years (Table 7). In fact, %IRI variability for C. gariepinus was significantly (p = 0.03) related to flooded area while that of S. intermedius was significantly (p = 0.01) related to discharge (Table 6).

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287

4. Discussion

288

Our results show that seasonal changes in the fish assemblage of the Okavango delta are highly correlated to hydrological variables. Discharge (flood pulse) was the strongest driver of seasonal change, while mean flooded area (contraction and expansion) had a lesser effect. These results are in accordance to the flood pulse concept by Junk et al. (1989) and consistent with observations from other floodplain systems (Lowe-McConnell, 1987; Arrington et al., 2005). The observed fish assemblage dynamics are driven by C. gariepinus, which dominates the experimental catches at minimum discharge (coinciding with maximum flooded area) and H. vittatus, which dominates during peak discharge (coinciding with small flooded area). The low discharge period interfaces with decreasing water levels where fish that were dispersed on the inundated floodplains back-migrate to the main channel and oxbow

289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

9

Table 3 Results of regression analysis between two hydrological variables and several fish indices of selected fish species from the Delta, where only the significant relationships (at 95% confidence level) are shown. Hydrological variable

Fish index

Species

DF

F

P

R2

Beta

Discharge

No/set

C. gariepinus S .intermedius C. ngamensis M. altisambesi All species C. gariepinus S. intermedius C. ngamensis All Species H. vittatus S. intermedius M. altisambesi All Species H. vittatus C. ngamensis M. altisambesi

11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11

12.27 9.58 9.84 10.41 16.2 16.66 9.89 13.02 21.72 6.43 6.46 7.43 17.09 13.45 9.65 11.34

0.006 0.011 0.011 0.009 0.002 0.002 0.01 0.005 0.001 0.03 0.029 0.021 0.002 0.004 0.011 0.007

0.55 0.49 0.5 0.51 0.62 0.62 0.5 0.57 0.68 0.39 0.39 0.43 0.63 0.57 0.49 0.53

0.74 0.7 0.7 0.71 0.79 0.79 0.71 0.75 0.83 0.63 0.63 0.65 0.79 0.76 0.7 0.73

C. gariepinus H. vittatus S. intermedius C. ngamensis All Species C. gariepinus H. vittatus S. intermedius C. ngamensis All Species S .intermedius C. gariepinus H. vittatus C. ngamensis

11 11 11 11 11 11 11 11 11 11 11 11 11 11

11.82 11.81 22.19 6.75 6.34 14.1 14.62 17.32 7.1 6.22 34.01 22.57 9.54 5.44

0.006 0.006 0.008 0.027 0.031 0.004 0.003 0.002 0.024 0.032 0 0.001 0.011 0.042

0.54 0.54 0.69 0.4 0.39 0.59 0.59 0.63 0.42 0.38 0.77 0.69 0.49 0.35

0.74 0.74 0.83 0.64 0.62 0.77 0.77 0.8 0.64 0.62 0.88 0.83 0.7 0.59

g/set

ML

%IRI

Flooded area

No/set

g/set

ML %IRI

Table 4 Summary of direct gradient analysis (CCA) results to establish fish distribution pattern and spawning behavior of selected fish species along an environmental gradient in the Delta’s Panhandle. Correlation values in bold indicate the key environmental variable in that particular ordination axis. The top four key species are placed in order of importance, starting with those having the highest scores under each axis. Spawning dynamics

Seasonal variability Axis 1

0.105 Eigenvalue % variance explained 34.700 Pearson correlation 0.993 Species – environment correlations Discharge 0.736 Area 0.835 Temperature 0.698 DOC 0.876 pH 0.476 DO 0.008 Monte Carlo 0.002 randomization test (p value) of eigenvalues O. macrochir Key species S. altus H. vittatus S. codringtonii

306 307 308 309 310 311 312

Axis 2

Axis 3

0.067 22.100 0.896

0.038 12.500 0.962

0.323 0.274 0.421 0.139 0.163 0.419 –

0.497 0.099 0.429 0.111 0.623 0.063 –

S. S. S. S.

C. rendalli H. cuvieri S. intermedius C. gariepinus

giardi macrocephalus thumbergi carlottae

lagoons. The contraction period (decreasing inundation) results in increased piscivory (Mosepele et al., 2009), where small prey species like M. altisambesi become vulnerable to predation from C. gariepinus (Merron, 1993). The dominance of the fish assemblage by C. gariepinus during the contraction period is perhaps an effect of what Merron (1993) calls the catfish run. Increased activity by

Axis 1 Eigenvalue 0.102 % variance explained 45.100 Pearson correlation 0.993 Species – environment correlations Discharge 0.750 Area 0.639 Temp 0.634 DOC 0.728 pH 0.481 DO 0.274 Monte Carlo 0.004 randomization test (p value) of eigenvalues Key Species H. vittatus S. nigromaculatus C. gariepinus

Axis 2

Axis 3

0.051 22.500 0.981

0.023 10.100 0.988

0.455 0.377 0.758 0.207 0.142 0.434

0.307 0.636 0.094 0.576 0.12 0.495

O. macrochir C. rendalli S. woosnami S. leopardinus

S. robustus O. andersonii S. angusticeps H .cuvieri

Clarias during this period (Merron, 1993) is reflected in a higher dominating catch rate in the experimental nets. Conversely, H. vittatus is an open water predator which is found predominantly in perennial water habitats (Winemiller and Kelso-Winemiller, 1994). High seasonal fluctuations in the fish assemblage structure (Table 2) are possibly caused by longitudinal migrations (Økland et al., 2005)

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

313 314 315 316 317 318 319

G Model

ECOHYD 132 1–14 10

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Fig. 6. Cluster analysis illustrating (A) seasonal and (B) annual fish assemblage patterns along a hydrological gradient.

320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343

during increasing and decreasing water levels. There were no significant changes in the density and relative abundance of H. vittatus between peak and minimum discharge periods (Fig. 4). Our observations confirm previous findings (Winemiller and Kelso-Winemiller, 1994; Merron and Bruton, 1995) that the local distribution of H. vittatus is regulated by the flood regime. H. vittatus is primarily a resident channel species (Merron, 1991), where it dominates the fish assemblage at peak discharge. Floodplains are characterized by shifting mosaics (Ward and Tockner, 2001), driven by dilution (discharge), expansion (flooded area) and contractions (drawdown) which is reflected in the fish population dynamics observed in this study. During maximum inundation (July–September) the habitat is more homogeneous where channels, lagoon and oxbows are connected to create one continuous landscape mosaic, reflected in strong seasonal similarity (Fig. 4). This period also corresponds to high species (a) diversity in the Delta (Fig. 3). The low fish assemblage similarity observed between April and August is probably caused by different dispersion and breeding behavior during increasing water levels and subsequent inundation. The flood pulse starting from Mohembo results initially in water level increases within the main channel

only (hence no marked changes in habitat or a diversity) and only longitudinal fish movements. When the river channel is full, water spills over onto the peripheral floodplains and facilitates lateral fish migrations (feeding and breeding) which would then cause variations in a diversity. Flood pulse dynamics in the Delta are characterized by a 3–4 month time lag between peak flooding at Mohembo and maximum flooded area expansion (Wolski et al., 2005; Fig. 1) and reflected in the associations with the two hydrological variables as illustrated in Figs. 4 and 5. Generally, there is high dissolved oxygen (DO) deficiency in floodplain systems at drawdown periods when decomposition processes outstrip production (Bayley, 1995) but localized hypoxia in the Delta also occurs when anoxic water is flushed out from beneath papyrus beds by new floods (Sethebe, 2011). These two processes explain the two low DO periods observed in the Delta during April and October/November (Fig. 1). Our results show that DO had a stronger association with species distribution during the latter period (Fig. 5). Two species, S. macrostoma (a mochokid) and Oreochromis andersonii (a cichlid) were highly associated with the DO gradient (i.e. DO) and have tolerance for hypoxic conditions (Chapman et al., 1994).

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

11

Table 5 Summary of spawning season of selected species from the Okavango Delta with results of regression analysis where values in the table are partial p values for each independent environmental factor. P values in bold indicate the environmental factor with the strongest effect while p values with an * indicate a significant relationship at the 95% confidence level.

368 369 370 371 372

Species

Spawning season

Peak month

Discharge

Area

Temp

pH

DO

R2

S. carlottae S. codringtonii S. macrocephalus C. ngamensis P. okavangensis M. altisambesi H. cuvieiri T. sparrmanii S. angusticeps P. acuticeps B. lateralis M. lacerda C. gariepinus P. ngamensis O. andersonii S. intermedius S. nigromaculatus S. vanderwaali S. thamalakanensis S. woosnami B. poechii S. macrostigma C. rendalli S. leopardinus H. vittatus

Aug–Nov, Jan–Feb Aug–Nov, Jan–Feb August–January August–March December–March December–April July–March Jun, Aug, Oct–Dec June–January June, August–October May–November Nov– March, August November–June November, January Oct–Nov, Jan–Apr, Jun October–February October–March October–March October–March October, Jan–February Sep–Dec, August, May Sep–Oct, Jan–February Sep–Jan, March, May September–March September–April

August February September, October January February January January, November October January June August August January January January January February February January, October January September January, October October January January

0.00* 0.05 0.04* – – 0.24 0.00* 0.04* 0.00* – – 0.44 0.02* 0.01* – 0.00* – – – – 0.99 0.24 – – 0.01*

– – 0.11 – 0.01* 0.01* 0.00* – 0.03* – (0.01)* 0.53 0.00* 0.03* 0.047* 0.00* – – – – 0.00* – 0.03* – (0.00)*

– 0.06 0.01* – – – – – 0.15 0.08 – 0.83 – – – – 0.07 0.01* 0.04* 0.19 0.00* 0.11 0.01* 0.00* –

– 0.14 – 0.09 – – – 0.27 – – – 0.48 – – – – 0.30 – – 0.33 – – – – –

– 0.30 0.20 0.01* 0.04* 0.33 0.03* 0.16 – – 0.23 0.75 – 0.28 – – 0.05 0.09 0.14 0.02* 0.13 0.26 – 0.01* –

0.67 0.79 0.91 0.63 0.76 0.85 0.94 0.53 0.73 0.30 0.62 0.20 0.85 0.73 0.37 0.83 0.73 0.69 0.55 0.72 0.90 0.54 0.57 0.82 0.80

This confirmed van der Waal’s (1998) observation that silurids are not structured along oxygen gradients (see Fig. 5) due to vestigial lungs which allows them to breathe atmospheric oxygen (Lowe-McConnell, 1987). Spawning season for most of the species in this study was also

Fig. 7. Results of DCA/DECORANA analysis illustrating inter-annual patterns in fish assemblage shifts in the study area.

correlated to the hydrological mediated environmental gradient. The high water season in tropical (floodplain) systems creates new micro-habitats, greatly expands the living space, enhances food availability for juvenile fish, and decreases inter specific competition for food and other resources (Lowe-McConnell, 1987; Chapman and Frank, 2000). Accordingly, this is the period when most floodplain fish species migrate into the marshes for feeding and reproduction (Chapman and Chapman, 2003). Results from our study are consistent with this pattern. CCA revealed that while the onset of spawning of individual species is lagged, and possibly cued by different environmental factors (Fig. 5), the majority is tied to mean flooded area (Table 5). When water levels rise, terrestrial vegetation is submerged and nutrients leaching from decomposing organic matter (dung, terrestrial grass, shrubs and trees), and directly from inflowing water, result in increased plankton production (McLachlan, 1970, Q3 1974, Kolding, 1993), which will cascade into fish production. Similar to Chapman and Chapman’s (2003) observations, this study also shows that some fish species have discrete breeding seasons that correspond to seasonal flooding (e.g. P. okavangensis); other species are multiple spawners that spawn at the onset of the floods until peak flooding (e.g. M. altisambesi); while other species spawn throughout the year, with strong seasonal peaks (e.g. O. andersonii, a mouth-brooder). Mouth-brooders enhance survivor-ship as the fry is protected from predation and ventilated to supply sufficient oxygen, which allows them to spawn throughout the year (Corrie et al., 2008). Earlier studies have aggregated fish spawning behavior into a single event driven by either flooding (Duque et al., 1998), oxygen (Chapman and Frank, 2000), temperature (Humphries et al., 1999), or inundated area (Halls et al., 1999). But

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

12

Table 6 Summary of regression analysis between the two hydrological variables and some indices of selected fish species from the Delta based on an annual scale. Significant values (at 95% level) are in bold. Hydrological Factor

Species

Index

DF

F

p

R2

Beta

Area Area Area Area Area Discharge Discharge Discharge Discharge Discharge Discharge

C. gariepinus S. intermedius M. altisambesi M. altisambesi M. altisambesi S. intermedius S. intermedius S. intermedius C. ngamensis H. vittatus M. altisambesi

%IRI no/set %IRI g/set no/set %IRI g/set no/set g/set no/set g/set

7 7 7 7 7 7 7 7 7 7 7

7.88 4.91 4.57 3.94 6.77 11.6 39.37 17.61 4.52 7.50 4.82

0.03 0.07 0.08 0.09 0.04 0.01 0.00 0.01 0.08 0.03 0.07

0.57 0.45 0.43 0.40 0.53 0.66 0.87 0.75 0.43 0.56 0.45

0.75 0.67 0.66 0.63 0.73 0.81 0.93 0.86 0.66 0.75 0.67

Table 7 Summary of assemblage stability values (CV) observed during the study period in the Delta of the top four species in the assemblage. Dominant species column refers to the most dominant species in the assemblage based on %IRI with values shown in [].

407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438

Year

CV

Description

Dominant species

2001 2002 2004 2005 2006 2007 2008 2009

78 77 87 73 90 101 73 70

Fluctuating Fluctuating Fluctuating Moderately fluctuating Fluctuating Fluctuating Moderately fluctuating Moderately fluctuating

C. gariepinus [32] C. gariepinus [31] S. intermedius [40] S. intermedius [37] C. gariepinus [30] S. intermedius [35] S. intermedius [33] S. intermedius [43]

there are also species that spawn during the low flow period (and high temperatures) like S. macrocephalus, Synodontis thamalakanensis and Synodontis vanderwaali (Table 5). We can conclude that different species spawn at different times and different time spans, cued by different factors which all are likely driven by the hydrological regime. Possibly, these diverse spawning patterns as caused by niche differentiation that reduces inter-specific competition. According to Kramer (1978), diversity in spawning patterns of tropical freshwater fish can be explained by five hypotheses; spawning is controlled by (i) adult and or juvenile food availability, (ii) inter-specific competition for food among juveniles, (iii) competition for spawning sites, (iv) a mechanism for reproductive isolation, and (v) is unrelated to local conditions but is a reflection of phylogenic specialization under particular conditions. While we did not specifically test these hypotheses, our study revealed diverse reproductive strategies that appear consistent with Kramer’s hypotheses. As rising water levels successively connects the habitats from channels, through lagoons and oxbows to the floodplains, we can tentatively suggest three kinds of reproductive strategies that describe fish spawning in the Delta; (I) channel + lagoon spawners whose spawning season peaks between January and March; (II) floodplain spawners whose spawning season peaks between June and October; and (III) channel + lagoon and floodplain spawners, whose spawning season peaks between (I) and (II) above. Generally, Silurids and Mochokids were strategy (I) species, Tilapia spp. were strategy (II) species, while Helicarion cuvieiri was a strategy (III) species.

4.1. Annual dynamics

439

S. intermedius and C. gariepinus dominated the Delta’s fish assemblage between 2001 and 2009. S. intermedius is a highly fecund, total spawner (Welcomme, 1985) while C. gariepinus is a multiple spawning, fast growing species (van der Waal, 1985). Multiple spawning requires higher annual reproductive effort than single spawning events (Burt et al., 1988), and enhances reproductive success (Cambray and Bruton, 1985). Lowe-McConnell (1987) observed that poor flood years normally resulted in recruitment failure of total spawners. This suggests that multiple spawning fish (such as C. gariepinus) are better competitors during poor flood years than total spawners (such as S. intermedius). Agostinho et al. (2000) observed that piscivorous fish species dominate floodplain assemblages during low/poor flood years, which suggest higher survival rates for predators when the fish assemblage is concentrated in the main channels. We argue that since C. gariepinus is a large sized predator that sometimes preys on the smaller S. intermedius (Winemiller and Winemiller, 1996; Mosepele et al., 2012), its relative survival increases significantly during low/poor flood years. Moreover, C. gariepinus has vestigial breathing apparatus which allows it to tolerate adverse environmental conditions (van der Waal, 1998). S. intermedius is an opportunistic species with small egg size and high fecundity that takes advantage of optimum conditions (Montcho et al., 2011). In conclusion, we observed that C. gariepinus dominated the fish assemblage during low flood year years, while S intermedius dominated the fish assemblage during high flood years. This is in agreement to van der Waal’s (1998)

440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491

observations that C. gariepinus usually dominate fish communities at low water levels, while high flood years appears more dominated by S. intermedius. High flood years are normally associated with fast fish growth rates (Dudley, 1974) where fish populations are normally comprised of survivors of the previous year and young-of-year fish (de Graaf, 2003). Therefore, high flood years normally result in a ‘‘boom’’ in fish production, followed by a ‘‘bust’’ during years of low floods (Arthington and Balcombe, 2011; Rayner et al., 2015). We see S. intermedius as a typical representative of this phenomenon, being highly fecund (Merron and Mann, 1995; Table 1) and fast growing (Mosepele and Nengu, 2003). The high flood years (2004, 2007–2009) therefore created optimum conditions for ‘‘booms’’ in S. intermedius populations, while low flood years (2001, 2002) resulted in a ‘‘bust’’. The residual effect of the high flood years was observed in 2005 where S intermedius still dominated the fish assemblage, though this was a low flood year. This observation is consistent with Welcomme (1985) who highlighted the variable effect of flooding dynamics on fish growth and yield in the Senegal, Niger and Logone rivers.

492

5. Conclusion

493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515

Our study has shown that the fish dynamics in the Okavango Delta is strongly influenced by the hydrological regime, which is in accordance with Junk et al.’s (1989) flood pulse concept. However, different fish species in the Delta responded differently to variations in discharge, flooded area and water chemistry. Seasonal dynamics appeared more pronounced than inter-annual fluctuations, but overall higher flooding resulted in higher fish production. One major conclusion from this study is that any future hydrological developments, such as abstraction or regulation in the Delta should take into account the effects on the fish assemblages. Consistent low flow regimes resulting from flow releases from upstream developments (e.g. dams in Angola and Namibia) will result in lower fish production and a fish assemblage dominated by large resilient species such C gariepinus. Conversely, high flood regimes above average will result in higher fish production and a fish assemblage dominated by smaller opportunistic species such as S. intermedius. Similar changes will be observed if the flooding regime is altered by future climate change scenarios with either higher or lower precipitation and ensuing flood levels.

516 Q5 Uncited reference 517 518

Maes et al. (2004). Conflict of interest

519

None declared.

520

Ethical statement

521 522

Authors state that the research was conducted according to ethical standards.

13

Acknowledgements None.

523 524

Funding body None.

525 526

References

527

Agostinho, A.A., Thomaz, S.M., Minte-Vera, C.V., Winemiller, K.O., 2000. Biodiversity in the high Parana´ River floodplain. In: Gopal, B., Junk, W.J., Davis, J.A. (Eds.), Biodiversity in Wetlands: Assessment, Function and Conservation, vol. 1. Backhuys Publishers, The Netherlands, pp. 89–118. Arrington, D.A., Winemiller, K.O., Layman, C.A., 2005. Community assembly at the patch scale in a species rich tropical river. Oecologia 144, 157–167. Arthington, A.H., Balcombe, S.R., 2011. Extreme flow variability and the ‘boom and bust’ ecology of fish in arid-zone floodplain rivers: a case history with implications for environmental flows, conservation and management. Ecohydrology 4, 708–720. Bayley, P.B., 1995. Understanding large river: floodplain systems. Bioscience 45 (3), 153–158. Burt, A., Kramer, D.L., Nakatsuru, K., Spry, C., 1988. The tempo of reproduction in Hyphessobrycon pulchripinnis (Characidae), with a discussion on the biology of ‘multiple spawning’ in fishes. Environ. Biol. Fishes 22 (1), 15–27. Cambray, J.A., Bruton, M.N., 1985. Age and growth of a colonizing minnow, Barbus anoplus, in a man-made lake in South Africa. Environ. Biol. Fishes 12 (2), 131–141. Cantu, N.E.V., Winemiller, K.O., 1997. Structure and habitat associations of Devils River fish assemblages. Southwest Nat. 42 (3), 265–278. Chapman, L.J., Chapman, C.A., 2003. Fishes of the African rain forests: emerging and potential threats to a little known fauna. In: Crisman, T.L., Chapman, L.J., Chapman, C.A., Kaufman, L.S. (Eds.), Conservation, Ecology and Management of African Fresh Waters. University Press of Florida, USA, pp. 176–209. Chapman, L.J., Kaufman, L., Chapman, C.A., 1994. Why swim upside down? A comparative study of two Mochokid catfishes. Copeia 1, 130–135. Chapman, L.J., Frank, C.E., 2000. Breeding seasonality in the African cyprinid Barbus neumayeri. Afr. J. Trop. Hydrobiol. Fish. 9, 36–48. Clarke, K.R., Gorley, R.N., 2001. PRIMER v5 (& v6): User Manual/tutorial. PRIMER-E. Plymouth, United Kingdom. Corrie, L.W., Chapman, L.J., Reardon, E.E., 2008. Brood protection at a cost: mouth brooding under hypoxia in an African cichlid. Environ. Biol. Fishes 82, 41–49. de Graaf, G., 2003. The flood pulse and growth of floodplain fish in Bangladesh. Fish. Manage. Ecol. 10, 241–247. Dudley, R.G., 1974. Growth of Tilapia of the Kafue Floodplain, Zambia: predicted effects of the Kafue Gorge Dam. Trans. Am. Fish. Soc. 2, 281– 291. Duque, A.B., Taphorn, D.C., Winemiller, K.O., 1998. Ecology of the coporo, Prochilodus mariae (Characiformes. Prochilondontidae), and status of annual migrations in western Venezuela. Environ. Biol. Fishes 53, 33–46. Freeman, M.C., Crawford, M.K., Barrett, J.C., Facey, D.E., Flood, M.G., Hill, J., Stouder, D.J., Grossman, G.D., 1988. Fish assemblage stability in a southern Appalachian stream. Can. J. Fish. Aquat. Sci. 45, 1949–1958. Gondwe, M., Masamba, W.R.L., 2013. Spatial and temporal dynamics of diffusive methane emissions in the Okavango Delta, Northern Botswana, Africa. Wetl. Ecol. Manage., http://dx.doi.org/10.1007/s11273013-9323-5. Grossman, G.D., Dowd, J.F., Crawford, M., 1990. Assemblage stability in stream fishes: a review. Environ. Manage. 14 (5), 661–671. Halls, A.S., Hoggarth, D.D., Debnath, K., 1999. Impacts of hydraulic engineering on the dynamics and production potential of floodplain fish populations in Bangladesh. Fish. Manage. Ecol. 6, 261–285. Hart, R.K., Calver, M.C., Dickman, C.R., 2002. The index of relative importance: an alternative approach to reducing bias in descriptive studies of animal diets. Wildlife Res. 29, 415–421. Humphries, P., King, A.J., Koehn, J.D., 1999. Fish, flows and flood plains: links between freshwater fishes and their environment in the Murray-Darling River system, Australia. Environ. Biol. Fishes 56, 129–151. Ibarra, M., Stewart, D.J., 1989. Longitudinal zonation of sandy beach fishes in the Napo River Basin, Eastern Ecuador. Copeia 2, 364–381.

528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

G Model

ECOHYD 132 1–14 14

596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659

K. Mosepele et al. / Ecohydrology & Hydrobiology xxx (2016) xxx–xxx

Junk, W.J., Bayley, P.B., Sparks, R.E., 1989. The flood pulse concept in riverfloodplain systems. Can. Spec. Publ. Fish. Aquat. Sci. 106, 110–127. Kolding, J., 1993. Population dynamics and life-history styles of Nile tilapia, Oreochromis niloticus, in Ferguson’s Gulf, Lake Turkana, Kenya. Environ. Biol. Fishes 37 (1), 25–46. Kolding, J., Asmund, A., 2010. Pasgear II, Version 2.3. University of Bergen, Norway. Kramer, D.L., 1978. Reproductive seasonality in the fishes of a tropical stream. Ecology 59 (5), 976–985. Lowe-McConnell, R.H., 1987. Ecological Studies in Tropical Fish Communities. Cambridge University Press, United Kingdom. Maes, J., Damme, S.V., Meire, P., Ollevier, F., 2004. Statistical modelling and environmental influences on the population dynamics of an estuarine fish community. Mar. Biol. 145, 1033–1042. McCarthy, T.S., Ellery, W.N., Bloem, A., 1998. Some observations on the geomorphological impact of hippopotamus (Hippopotamus amphibius L.) in the Okavango Delta, Botswana. Afr. J. Ecol. 36, 44–56. McCune, B., Mefford, M.J., 2006. PC-ORD, Multivariate Analysis of Ecological Data. Version 5.10 MjM Software. Gleneden Beach, Oregon, U.S.A. McCune, B., Mefford, M.J., 1999. Multivariate Analysis of Ecological Data, Version 4.24. Gleneden Beach, OR, USA. McLachlan, A.J., 1970. Some effects of annual fluctuations in water level on the larval chironomid communities of Lake Kariba. J. Anim. Ecol. 39, 79–90. Merron, G.S., 1991. The ecology and management of the fishes of the Okavango Delta, Botswana, with particular reference to the role of the seasonal floods. (PhD thesis)Rhodes University, South Africa. Merron, G.S., 1993. Pack-hunting in two species of catfish, Clarias gariepinus and C. ngamensis in the Okavango Delta, Botswana. J. Fish Biol. 43 (4), 575–584. Merron, G.S., Bruton, M.N., 1988. The ecology and management of the fishes of the Okavango delta, Botswana, with special reference to the role of the seasonal floods. J.L.B. Smith Institute of Ichthyology, Grahamstown, South Africa. Investigational Report No. 29 291 p. Merron, G.S., Bruton, M.N., 1995. Community ecology and conservation of the fishes of the Okavango Delta, Botswana. Environ. Biol. Fishes 43, 109–119. Merron, G.S., Mann, B.Q., 1995. The reproductive and feeding biology of Schilbe intermedius Ruppell in the Okavango Delta, Botswana. Hydrobiologia 308, 121–129. Mendelsohn, J.M., VanderPost, C., Ramberg, L., Murray-Hudson, M., Wolski, P., Mosepele, K., 2010. Okavango Delta: Floods of Life. RAISON, Namibia. Minns, C.K., 1989. Factors affecting fish species richness in Ontario lakes. Trans. Am. Fish. Soc. 118 (5), 533–545. Montcho, S.A., Chikou, A., Lale`ye`, P.A., Linsenmair, K.E., 2011. Population structure and reproductive biology of Schilbe intermedius (Teleostei: Schilbeidae) in the Pendjari River, Benin. Afr. J. Aquat. Sci. 36 (2), 139– 145. Mosepele, K., 2000. Length based fish stock assessment of the main exploited fish stocks of the Okavango delta, Botswana. (MPhil Thesis)University of Bergen, Norway. Mosepele, K., Nengu, S., 2003. Growth, mortality, maturity and lengthweight parameters of selected fishes from the Okavango Delta, Botswana. In: Pauly, D., Diouf, B., Samb, B., Vakily, M.J., Palamores, M.L.D. (Eds.), Fish Biodiversity: Local Studies as Basis for Global Inferences. Brussels, Belgium, pp. 67–74. Mosepele, K., Mosepele, B., Wolski, P., Kolding, J., 2012. Dynamics of the feeding ecology of selected fish species from the Okavango River delta, Botswana. Acta Ichthyol. Piscat. 42 (4), 271–289. Mosepele, K., Moyle, P.B., Merron, G.S., Purkey, D., Mosepele, B., 2009. Fish, floods and ecosystem engineers; interactions and conservation in the Okavango Delta, Botswana. Bioscience 59 (1), 53–64.

Nikolsky, G.V., 1969. Theory of Fish Population Dynamics as the Biological Background for Rational Exploitation and Management of Fishery Resources. Olivier and Boyd, United Kingdom. Oberdoff, T., Porcher, J., 1992. Fish assemblage structure in Brittany streams (France). Aquat. Living Resour. 5, 215–223. Økland, F., Thorstad, E.B., Hay, C.J., Næsje, T.F., Chanda, B., 2005. Patterns of movement and habitat use by tigerfish (Hydrocynus vittatus) in the upper Zambezi River (Namibia). Ecol. Freshw. Fish 14, 79–86. Ramberg, L., Wolski, P., Krah, M., 2006a. Water balance and infiltration in a seasonal floodplain in the Okavango Delta, Botswana. Wetlands 26 (3), 677–690. Ramberg, L., Hancock, P., Lindholm, M., Meyer, T., Ringrose, S., Sliva, J., van As, J., VanderPost, C., 2006b. Species diversity of the Okavango Delta, Botswana. Aquat. Sci. 68, 310–337. Rayner, T.S., Kingsford, R.T., Suthers, I.M., Cruz, D.O., 2015. Regulated recruitment: native and alien fish responses to widespread floodplain inundation in the Macquarie Marshes, arid Australia. Ecohydrology 8, 148–159. Sethebe, K., 2011. Fish dynamics and water quality in the Okavango Delta, Botswana: The case of Guma lagoon. (MSc Thesis)University of The Free State, South Africa. Skelton, P.H., 2001. A Complete Guide to Freshwater Fishes of Southern Africa. Struik Publishers, Cape Town. Smith, P.A., 1976. An outline of the vegetation of the Okavango drainage system. In: Proceedings of the Symposium on the Okavango Delta and its future utilisation, Botswana Society, Gaborone, pp. 93–112. ter Braak, C.J.E., 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67, 1167–1179. Tockner, K., Malard, F., Ward, J.V., 2000. An extension of the flood pulse concept. Hydrol. Process. 14, 2861–2883. van der Waal, B.C.W., 1985. Aspects of the biology of larger fish species of Lake Liambezi, Caprivi, South West Africa. Madoqua 14 (2), 101–1441. van der Waal, B.C.W., 1998. Survival strategies of sharptooth catfish Clarias gariepinus in desiccating pans in the northern Kruger National Park. KOEDOE 41 (2), 131–138. Ward, J.V., Tockner, K., 2001. Biodiversity: towards a unifying theme for river ecology. Freshw. Biol. 46, 807–819. Welcomme, R.L., Brummett, R.W., Denny, P., Hasan, M.R., Kaggwa, R.C., Kipkemboi, J., Mattson, N.S., Sugunan, V.V., Vass, K.K., 2006. Water management and wise use of wetlands: enhancing productivity. Ecol. Stud. 190, 155–180. Welcomme, R.L., 1985. River Fisheries. FAO Fisheries Technical Paper No. 262. FAO, Rome. Welcomme, R.L., 2007. Conservation of fish and fisheries in large river systems. Am. Fish. Soc. Sympos. 49, 587–599. Winemiller, K.O., Kelso-Winemiller, L.C., 1994. Comparative ecology of the African pike, Hepsetus odoe, and tiger-fish. Hydrocynus forskahlii, in the Zambezi River floodplain. J. Fish Biol. 45 (2), 211–225. Winemiller, K.O., Winemiller, L.C., 1996. Comparative ecology of catfishes of the Upper Zambezi River floodplain. J. Fish Biol. 49, 1043–1061. Wolski, P., Masaka, T., Raditsebe, L., Murray-Hudson, M., 2005. Aspects of dynamics of flooding in the Okavango delta. Botswana Notes Rec. 37, 179–195. Wolski, P., Savenije, H.H.G., 2006. Dynamics of floodplain-island groundwater flow in the Okavango Delta, Botswana. J. Hydrol. 320, 283–301. Wolski, P., Savenije, H.H.G., Hudson, M.M., Gumbricht, T., 2006. Modelling of the flooding in the Okavango Delta, Botswana, using a hybrid reservoir-GIS model. J. Hydrol. 331, 58–72. Zeug, S.C., Winemiller, K.O., 2007. Ecological correlates of fish reproductive activity in floodplain rivers: a life-history-based approach. Can. J. Fish. Aquat. Sci. 64, 1291–1301.

Please cite this article in press as: Mosepele, K., et al., Fish community dynamics in an inland floodplain system of the Okavango Delta, Botswana. Ecohydrol. Hydrobiol. (2017), http://dx.doi.org/10.1016/j.ecohyd.2017.01.005

660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722